Create a Data Quality Audit Protocol

You are a senior MEAL specialist with expertise in data quality assurance. Your task is to create a comprehensive Data Quality Audit (DQA) Protocol for a program. **Program Context:** - Program name: the program requiring data quality assessment - Number of implementation sites: the program's geographic footprint - Key indicators to audit: the priority indicators for verification - Data collection tools: how data is currently collected - Reporting frequency: how often data is reported - Donor: the primary funding agency **Deliverables:** **1. Audit Objectives and Scope** - Purpose of the DQA (verification, improvement, compliance, or all three) - Specific data quality dimensions to assess: Validity, Reliability, Completeness, Timeliness, Precision, and Integrity (define each in the program context) - Time period under review - Data sources and levels to audit (facility, district, national) **2. Sampling Strategy** - Sampling frame: list of all sites and reporting units - Sample size calculation: recommend auditing at least 10-15% of sites, with justification - Selection method: random sampling with stratification by performance level, geography, or facility type - Indicator sampling: which indicators to audit in full versus spot-check - Document sampling: how many records to verify per site (recommend minimum 20 records or 10% of total, whichever is larger) **3. Verification Procedures** For each data quality dimension, provide specific verification steps: | Dimension | Verification Method | Data Source Cross-Reference | Pass Criteria | |---|---|---|---| Include: - **Source document verification:** Compare reported figures against original source documents (registers, forms) - **Recounting:** Independently recount a sample of records and compare with reported totals - **Cross-system checks:** Compare figures across parallel systems (e.g., facility register vs. digital form vs. monthly report) - **Timeliness check:** Percentage of reports submitted on time - **Completeness check:** Percentage of required fields completed in source documents - **Logic checks:** Identify impossible or implausible values **4. Scoring Rubric** Create a scoring matrix for each dimension on a 1-4 scale: | Score | Label | Description | |---|---|---| | 4 | Highly Satisfactory | Data meet all quality standards with minor issues only | | 3 | Satisfactory | Data are generally reliable with some correctable gaps | | 2 | Needs Improvement | Significant quality issues that affect data reliability | | 1 | Unsatisfactory | Data cannot be relied upon for decision-making | Define what constitutes each score for each dimension. Provide an overall composite scoring method with weighting. **5. Audit Tools and Templates** List and briefly describe each tool needed: - Site-level verification checklist - Record recounting worksheet - Data cross-reference matrix - Timeliness tracking log - Findings summary template - Corrective action plan template **6. Corrective Action Framework** For each score level (1-4), specify: - Required corrective actions - Responsible party - Timeline for remediation - Follow-up verification date - Escalation protocol if issues persist **7. Audit Schedule and Team** - Recommended audit frequency (quarterly for high-risk indicators, semi-annually for routine) - Team composition and roles - Estimated level of effort per site (person-days) - Budget considerations Align with USAID DQA guidelines, Global Fund LFA verification standards, and MEASURE Evaluation data quality frameworks.